You are better than me!

The ongoing frontier

Over the past few months, you have seen frontier models make more and more progress. Opus 4.5 was a big deal at the end of 2025, in its own league at the time. Now Opus 5.5, Astra 6, and Argon are very solid, not to mention SpaceX catching up with Grok and the whole open-weight frontier initiative being primarily driven by DeepSeek.

The frontier is expanding in both depth and breadth, and it won’t stop!

Where are the humans?

I guess we are all starting to feel a bit of fear. What’s happening? Is AI going to replace us? What will happen to my job? Is my job replaceable? How will I pay my bills? These are all human emotions. Because of that there has been a negative frenzy driven by pure pessimism and a doomer approach to human life and action, mostly by people who can’t handle their emotions and egos. And yet, we are the ones improving the frontier!

You are better than me

As an engineer, I have a distinct approach to this: “You are better than me, you write better code!” You should admit this to yourself as soon as possible. The reason is that there will always be plenty of problems to solve, since human curiosity is infinite.

As David Deutsch puts it, problems are inevitable, but problems are also soluble.

I’d much rather solve bigger and more problems than spend days writing small patches of code. Once we have self-replicating AI systems, the world will be in an ideal place. That may take some time, though, so we’d better roll up our sleeves.

Typically, I write when something nudges me to communicate with myself and then with others. This time, I have David Heinemeier Hansson to thank. He has been quite vocal about the impact of this on the industry, programming languages, and people, as I wrote before in my post on programming languages. So the time has come to officially wave a white flag and admit to LLM:

“You are better than me. Now improve that architecture and don’t make mistakes!”